US2357014A - Flux composition - Google Patents

Description

Patented Aug. 29, 1944 FLUX COMPOSITIOR Menahem Merlub-Sobel, Jersey City, N. J., and

Jerome M. Bialosky, Portsmouth, Ohio, assignors to William Ohio L. Ulmer, Cleveland Heights,

No Drawing. Application July 1, 1941,

Serial No. 401,332

6 Claims.

This invention relates to fluxes for brazing and welding, and, in particular, to fluxes for use in such brazing and welding operations as have required, in the past, high concentrations of corrosive fluorides (speciflcally acid fluorides) in the fluxing medium. Brazing with silver solders comes, especially, within the scope of our invention, both when such silver solder is used because of its relatively low melting point and when it is used because of its .capacity for joining and brazing such refractory or oxidizable metals as the chromium-iron alloys, of which stainless steel is the most common example.

It has long been known that, for silver solders and alloys of similar composition, it is necessary which integrates into K2B4O7+2KHF2+5HzO=4H3BO3+4KF It is particularly worthy of notice that no borofluoride is formed under these circumstances, and equally important that neutral potassium fluoride is formed in equimolecular quantity to the boric acid; this is of interest because, though the neutral fluorides have some minor fluxing qualities, they are far inferior to acid fluorides or even to boric acid and the borates in fluxing power.

If, as in some formulas proposed for such soldering and welding, the molecular ratio of bifluoride is more than twice the borate, then the excess bifluoride will react-but for the most partonly under the heat of the torch-in the following manner:

Once again, it should be noted that an excess of neutral alkali fluoride is formed, adding further to the content of this demi-inert constituent. Much more important is the fact that the formation of three mols of water for each mol of potassium fluoborate means that for each mol of this salt the heat of vaporization of three mols .01 water will be taken away from the very area where it is most needed. The resulting chilling oi the work right where and when heat is desirable must inevitably affect the quality of the weld adversely, especially since, to overcome this drawback, the heating torch must be kept on the work correspondingly longer. Quantitative slowing of production is thus added to the quality factors already mentioned.

Some formulas have proposed the hot mixing of borate solutions with the acid fluoride solutions, which will result, to some extent, in the formation of the alkali fluoborate prior to torch action. To the extent that the borate exceeds the proportions indicated by the reaction KsB407+14KHFz=12KF+4KBF4+7H2O (approximately 17.6% of borate, with the remainder bifluoride) the solution will be alkaline and this reaction will be impeded. If, on the other hand, the proper ratio is maintained-as is very seldom done-no free borate is available to the flux; this borate is very desirable because of its own fluxing action. In any event, there is always present, in chemically preponderant quantity, the relatively useless, and therefore undesirable, neutral fluoride. Some of this excess neutral fluoride can be avoidedand this procedure has been proposed in an earlier patent (Valentine, U. S. P. 2,099,- 582), by using as a flux a mixture of bifluoride and boric acid, which, despite excess boric acid in the proposed formula, will react, either by pre-action or under the torch, as follows:

This, however, is still open to grave objections. Unless the reaction is forced to completion, free acidity would remain, and thereby make for a material very corrosive in nature, not only to metals, but even to glass. This can, admittedly, be accomplished, but presents manufacturing dimculties. However, the neutral fluoride problem is by no means adequately solved; there are still two mols of weak and comparatively useless neutral alkali fluoride to every mol of fluxworthy potassium fluoborate, representing a two-thirds dilution. 1

We have found that, if, instead of attempting to form the fluoborate in situ during torch operations, this compound is used as such, either wholly, or at least significantly more so than possible hitherto, free from neutral fluorides, a

reactive with metallic oxides, just as silicon fluoride is known to be.

However, even simple fluoboratesalts are not completely satisfactory. Under the welding torch, for example, potassium fluoborate fumes badly, and does not wet the work as well as might be desired. Since wetting is an essential factor in the physics-chemistry of fluxing. it is obvious .that such fluoborate therefore does not reach the ideal as regard flux requirements. This, too, will explain why, while fluoborates have occasionally been suggested as desirable additional and minor constituents in' fluxes for silver solders and the like, they have been out of the question as the predominant factor in a commercial flux.

In our invention, we have found that, if, to a fluoborate composition containing comparatively little concomitant other fluorine compounds, there be added oxy-compounds of boron, such as boric acid or any of the alkali borates, there results a flux which has all the advantages of the fluoborate itself, such as non-corrosiveness, high fluxing activity, etc., yet free from the drawbacks of simple fluoborate, such as fuming and non-wetting. Moreover, the comparative absence of extraneous fluorides makes for high concentration of the active fluxin constituents. By comparative absence we refer to all compositions in which there is less than one mol of non-fluoborate fluoride to every mol of fluoborate. It will be recalled that none of the standard procedures yields mixtures with less than two mols of non-fluoborate fluorides to each mol of fluoborate.

As an example of a mixture acid in characterthough free from the corrosiveness inherent in compositions containing acid fluorides, we can cite a composition consisting of equal parts of potassium fluoborate'and of boric acid, with suflicient water to make a convenient paste.

In general, however, we view alkaline borates as preferable to boric acid, and, specifically, we have found most desirable those compositions in which the alkali content of the borate compound equals, or exceeds, that called for in the penta borate. The tetraborate, for example, makes an excellent medium for flux formulation. As one example of a, composition excellent as a flux, we may cite equal mixtures of potassium fluoborate and potassium tetraborate, again with suflicient water to make a convenient paste.

It is obvious that both the potassium fluoborate and the oxy-borat may be formed in solution, and the paste made in that way. For thefluoborate, it is absolutely essential that free hydrofluoboric acid first be formed by interaction of boric acid with hydrofluoric acid (not an acid fluoride, since this would give the undesirable neutral fluoride contaminant as a concomitant); the hydrofluoboric acid thus formed can then be neutralized by means of alkali carbonate or hydroxide. Alkali borate can similarly be of the resulting mixture of hydrofluoboric and boric acids by means of 32 grams of potassium hydroxide. The paste resulting from this reaction provides an excellent fluxing medium for silver solder, on both stainless steel and copperbase alloys.

If desired for some special reason, or if present as a contaminant, we have found that nonfluoborate fluorine compounds may be present to some extent without great damage. If, as already indicated, such non-fluoborate fluorides are present in an amount not exceeding mol for mol to the fluoborate content, little damage results, even though, in general, we prefer such extraneous fluorine compounds to be substantially absent.

In addition to alkali present as neutralization agent for the fluoboric acid and boric acid, we may optionally have present, in our fluxes, free bonate. Such free alkali has, in certain cases,

specific fluxing characteristics of its own. Such free alkali should not, however, exceed approximately forty per-cent of the total. As an example of a ternary mixture of this type, we may cite a flux whose solids contents consist 'of 40 parts by weight of potassium fluoborate, 40 parts of potassium pentaborate (hydrated), and 20 parts of potassium carbonate.

As limitations, we have found that the alkali fluoborate should always be no less than 20% of the total solids content of the flux, nor should it exceed 85% of the total. The oxy-compound of boron, either as boric acid or as an alkali boron compound, should be no less than 15% of the total solids, nor should it exceed Extraneous fluorine compounds, as previously stated, must under no circumstances exceed the equimolar quantity of the fluoborate present.

\ W claim:

borate more basic than the taborate.

pen

2. A flux containing from 20% to alkali fluoborate, and from 15% to 70% of an alkali borate more basic than the pentaborate.

3. A flux consisting mainly of an alkali flueborate and an alkali borate more basic than the pentaborate, any non-fluoborate fluorine compounds present being to an extent no greater than equimolar to the fluoborate.

4. A flux containing from 20% to 85% of an alkali fluoborate, and from 15% to 70% of an alkali borate more basic than the pentaborate,

any non-fluoborate fluorine compounds present being to an extent no greater than equimolar to the fluoborate.

5. A flux containing from 20% to 85% tassium fluoborate, and from 15% potassium borate.

6. A flux consisting mainly of potassium fluoof poto 70% oi borate and a potassium borate more basic than the pentaborate.

MENAHEM MERLUB-SOBEL. JEROME M. BIALOSKY.

of an,